Transcript UTP Cable Connectors

Biomedical Instrumentation
Prof. Dr. Nizamettin AYDIN
[email protected]
[email protected]
http://www.yildiz.edu.tr/~naydin
1
Biomedical Instrumentation
Lecture 3
Measurement Systems
2
Biomedical Instrumentation
• Diagnosis and therapy depend heavily on the
use of medical instrumentation.
• Medical procedures:
Medicine can be defined as a multistep
procedure on an individual by a physician,
group of physician, or an institute, repeated
until the symptoms disappear
3
The Importance of Biomedical Instrumentation
• Medical procedure
1 Collection of data - qualitative and/or
quantitative
2 Analysis of data
3 Decision making
4 Treatment planning based on the decision
4
Biomedical Instrumentation System
• All biomedical instruments must interface with
biological materials. That interface can by direct
contact or by indirect contact
5
Components of BM Instrumentation System…
• A sensor
– Detects biochemical, bioelectrical, or biophysical
parameters
– Provides a safe interface with biological materials
• An actuator
– Delivers external agents via direct or indirect
contact
– Controls biochemical, bioelectrical, or biophysical
parameters
– Provides a safe interface with biologic materials
6
…Components of BM Instrumentation System…
• The electronics interface
– Matches electrical characteristics of the
sensor/actuator with computation unit
– Preserves signal to noise ratio of sensor
– Preserves efficiency of actuator
– Preserves bandwidth (i.e., time response) of
sensor/actuator
– Provides a safe interface with the sensor/actuator
– Provides a safe interface with the computation unit
– Provides secondary signal processing functions for
the system
7
…Components of BM Instrumentation System
• The computation unit
–
–
–
–
provides primary user interface
provides primary control for the overall system
provides data storage for the system
provides primary signal processing functions for
the system
– maintains safe operation of the overall system
8
Problems Encountered in Measuring a Living System
• Many crucial variables in living systems are
inaccessible.
• Variables measured are seldom deterministic.
• Nearly all biomedical measurements depend
on the energy.
• Operation of instruments in the medical
environment imposes important additional
constraints.
9
The scientific method…
• In the scientific method, a hypothesis is tested
by experiment to determine its validity.
Problem
statement
Review
prior work
State
hypothesis
Perform
experiments
Design further
experiments
More
experiments
necessary
Analyze
data
Final
conclusions
Problem
solved
10
…The scientific method
• In the scientific method, a hypothesis is tested
by experiment to determine its validity.
– For example, we might hypothesize that exercise
reduces high blood pressure yet experimentation
and analysis are needed to support or refute the
hypothesis.
• Experiments are normally performed multiple times.
Then the results can be analyzed statistically to
determine the probability that the results might have
been produced by chance.
• Results are reported in scientific journals with enough
detail so that others can replicate the experiment to
confirm them.
11
Clinical diagnoses
• Physicians often need instrumentation to obtain
data as part of the scientific method.
– For example, a physician obtaining the history of a patient
with a complaint of poor vision would list diabetes as one
possibility on a differential diagnosis.
Chief
complaint
Obtain
history
List the
differential
diagnosis
Examination
and tests
Treatment
and
evaluation
Select further
tests
Use data
to narrow the
diagnosis
Final
diagnosis
More than
one likely
Only one
likely
12
Feedback in measurement systems…
• Figure shows that the measurand is measured by a
sensor converting the variable to an electrical signal,
which can undergo signal processing. Sometimes the
measurement system provides feedback through an
effector to the subject.
Outputs
Measurand
Sensor
Signal
conditioning
Feedback
Effector
Signal
processing
Data
storage
Data
displays
Data
communication
13
…Feedback in measurement systems
• Figure (a) shows that a patient reading an instrument usually
lacks sufficient knowledge to achieve the correct diagnosis.
• Figure (b) shows that by adding the clinician to form an effective
feedback system, the correct diagnosis and treatment result.
Patient
Patient
(a)
Instrument
Instrument
Clinician
(b)
14
…Feedback in measurement systems
• In certain circumstances, proper training of the patient and a
well-designed instrument can lead to self-monitoring and selfcontrol (one of the goals of bioinstrumentation).
– An example of such a situation is the day-to-day monitoring of glucose
by people suffering from diabetes. Such an individual will contact a
clinician if there is an alert from the monitoring instrument.
Clinician
Abnormal
readings
Patient
Instrument
15
Classifications of Biomedical Instruments
•
•
•
•
•
The sensed quantity
The principle of transduction
The organ system for measurement
The clinical medicine specialties
Based on the activities involved in the medical
care, medical instrumentation may be divided
into three categories:
– Diagnostic devices
– Therapeutic devices
– Monitoring devices
16
Generalized Medical Instrumentation System…
17
…Generalized Medical Instrumentation System…
• Measurand
– Physical quantity, property, or condition that the
system measures
•
•
•
•
•
•
•
•
Biopotantial
Pressure
Flow
Dimension (imaging)
Displacement (velocity, acceleration, and force)
Impedance
Temperature
Chemical concentrations
18
…Generalized Medical Instrumentation System…
Measurement
Range
Frequency, Hz
Method
Blood flow
1 to 300 mL/s
0 to 20
Electromagnetic or ultrasonic
Blood pressure
0 to 400 mmHg
0 to 50
Cuff or strain gage
Cardiac output
4 to 25 L/min
0 to 20
Fick, dye dilution
Electrocardiography
0.5 to 4 mV
0.05 to 150
Skin electrodes
Electroencephalography 5 to 300  V
0.5 to 150
Scalp electrodes
Electromyography
0.1 to 5 mV
0 to 10000
Needle electrodes
Electroretinography
0 to 900  V
0 to 50
Contact lens electrodes
pH
3 to 13 pH units
0 to 1
pH electrode
pCO2
40 to 100 mmHg
0 to 2
pCO2 electrode
pO2
30 to 100 mmHg
0 to 2
pO2 electrode
Pneumotachography
0 to 600 L/min
0 to 40
Pneumotachometer
Respiratory rate
2 to 50
breaths/min
0.1 to 10
Impedance
Temperature
32 to 40 °C
0 to 0.1
Thermistor
19
…Generalized Medical Instrumentation System…
• Sensor
– Converts a physical measurand to an electrical
output
• Should respond only to the form of energy present in the
measurand
• Should be minimally invasive (ideally noninvasive)
• Signal conditioning
– Amplify, filter, match the impedance of the sensor
to the display
– Convert analog signal to digital
– Process the signal
20
…Generalized Medical Instrumentation System…
• The specifications for a typical blood pressure sensor.
– Sensor specifications for blood pressure sensors are
determined by a committee composed of individuals from
academia, industry, hospitals, and government
Specification
Value
Pressure range
–30 to +300 mmHg
Overpressure without damage
–400 to +4000 mmHg
Maximum unbalance
±75 mmHg
Linearity and hysteresis
± 2% of reading or ± 1 mmHg
Risk current at 120 V
10 A
Defibrillator withstand
360 J into 50 
21
…Generalized Medical Instrumentation System…
• A hysteresis loop.
– The output curve obtained
when increasing the
measurand is different from
the output obtained when
decreasing the measurand.
Sensor
signal
Measurand
22
…Generalized Medical Instrumentation System…
• (a) A low-sensitivity sensor has low gain. (b) A
high sensitivity sensor has high gain.
Sensor
signal
Sensor
signal
Measurand
(a)
Measurand
(b)
23
…Generalized Medical Instrumentation System…
• Most sensors are analog and provide a continuous
range of amplitude values for output (a).
• Other sensors yield the digital output (b).
Amplitude
Amplitude
– Digital output has poorer resolution, but does not require
conversion before being input to digital computers and is
more immune to interference
Time
(a)
Time
(b)
24
…Generalized Medical Instrumentation System…
• Bioinstrumentation should be designed with a specific signal in
mind.
– Table shows a few specifications for an electrocardiograph
– The values of the specifications, which have been agreed upon by a
committee, are drawn from research, hospitals, industry, and government.
Specification
Value
Input signal dynamic range
±5 mV
Dc offset voltage
±300 mV
Slew rate
320 mV/s
Frequency response
0.05 to 150 Hz
Input impedance at 10 Hz
2.5 M
Dc lead current
0.1 A
Return time after lead switch
1s
Overload voltage without damage
5000 V
Risk current at 120 V
10 A
25
…Generalized Medical Instrumentation System…
(a) An input signal which exceeds the dynamic range.
(b) The resulting amplified signal is saturated at 1 V.
Amplitude
5 mV
(a)
Time
Dynamic
Range
-5 mV
Amplitude
(b)
1V
Time
-1 V
26
…Generalized Medical Instrumentation System…
• DC offset voltage is the amount a signal may be moved from
its baseline and still be amplified properly by the system.
Figure shows an input signal without (a) and with (b) offset.
Amplitude
(a)
Time
Amplitude
(b)
Dc offset
Time
27
…Generalized Medical Instrumentation System…
• The frequency response of a device is the range of frequencies
of a measurand that it can handle.
• Frequency response is usually plotted as gain versus frequency.
• Figure shows Frequency response of the electrocardiograph.
1.0
Amplitude
0.1
0.05 Hz
150 Hz
Frequency
28
…Generalized Medical Instrumentation System…
• Linearity is highly desirable for simplifying signal processing
(a) A linear system fits the equation y = mx + b. Note that all
variables are italic.
(b) A nonlinear system does not fit a straight line.
Output
Output
Input
(a)
Input
(b)
29
…Generalized Medical Instrumentation System…
Amplitude
Amplitude
• All bioinstrumentation observes the measurand either continuously or
periodically. However, computer-based systems require periodic
measurements since by their nature, computers can only accept discrete
numbers at discrete intervals of time.
(a) Continuous signals have values at every instant of time.
(b) Discrete-time signals are sampled periodically and do not provide
values between these sampling times.
Time
(a)
Time
(b)
30
…Generalized Medical Instrumentation System…
• Signal conditioning
– Amplify, filter, match the impedance of the sensor
to the display
– Convert analog signal to digital
– Process the signal
31
…Generalized Medical Instrumentation System…
• Output display
– Results must be displayed in a form that the human
operator can perceive
• Numerical, Graphical, Discrete or continuous,
Permanent or temporary, Visual or acoustical
• Auxiliary elements
–
–
–
–
Data storage
Data transmission
Control and feedback
Calibration signal
32
…Generalized Medical Instrumentation System…
• Panels and series
• Certain groups of measurements are often
ordered together because they are very
commonly used or because they are related.
• This may occur even if the measurements are
based on different principles or are taken with
different sensors.
• Table in next slide is an example of one of
these groups of measurements, which are
called panels or series.
33
…Generalized Medical Instrumentation System…
• Complete blood count for a male subject.
Laboratory test
Typical value
Hemoglobin
13.5 to 18 g/dL
Hematocrit
40 to 54%
Erythrocyte count
4.6 to 6.2  106/ L
Leukocyte count
4500 to 11000/ L
Differential count
Neutrophil 35 to 71%
Band 0 to 6%
Lymphocyte 1 to 10%
Monocyte 1 to 10%
Eosinophil 0 to 4%
Basophil 0 to 2%
34
…Generalized Medical Instrumentation System
• Hemoglobin is the protein which caries oxygen in the
bloodstream.
• Hematocrit is the percent of solid material in a given
volume of blood after it has been centrifuged.
• An erythrocyte is a red blood cell.
• A leukocyte is a white blood cell.
– The differential count tells how many of each type of white
blood cell there are in one microliter of blood.
– Unusual values for different leukocytes can be indicative of
the immune system fighting off foreign bodies.
35
Errors in measurements…
24Subat2k11
• When we measure a variable, we seek to determine the true
value, as shown in Figure (next slide) .
• This true value may be corrupted by a variety of errors. For
example
– Movement of electrodes on the skin may cause an undesired added
voltage called an artifact.
– Electric and magnetic fields from the power lines may couple into the
wires and cause an undesired added voltage called interference
– Thermal voltages in the amplifier semiconductor junctions may cause
an undesired added random voltage called noise. Temperature changes
in the amplifier electronic components may cause undesired slow
changes in voltage called drift.
• We must evaluate each of these error sources to determine their
size and what we can do to minimize them.
36
…Errors in measurements…
(a) Signals without noise are uncorrupted.
(b) Interference superimposed on signals causes
error.
Frequency filters can be used to reduce noise and
interference.
(a)
(b)
37
…Errors in measurements…
(a) Original waveform.
(b) An interfering input may shift the baseline.
(c) A modifying input may change the gain.
(a)
(b)
(c)
38
Accuracy and precision…
• Resolution
– the smallest incremental quantity that can be reliably measured.
• a voltmeter with a larger number of digits has a higher resolution than
one with fewer digits.
– However, high resolution does not imply high accuracy.
• Precision
– the quality of obtaining the same output from repeated
measurements from the same input under the same conditions.
– High resolution implies high precision.
• Repeatability
– the quality of obtaining the same output from repeated
measurements from the same input over a period of time.
39
…Accuracy and precision…
• Data points with
(a) low precision and (b) high precision.
40
…Accuracy and precision…
• Accuracy
– the difference between the true value and the measured
value divided by the true value.
• Obtaining the highest possible precision,
repeatability, and accuracy is a major goal in
bioinstrumentation design.
41
…Accuracy and precision…
• Data points with
(a) low accuracy and
(b) high accuracy
42
Calibration…
• Measuring instruments should be calibrated
against a standard that has an accuracy 3 to 10
times better than the desired calibration
accuracy.
• The accuracy of the standard should be
traceable to the institutions regulating the
standards (National Institute of Standards and
Technology, TSI, etc.) .
43
Calibration…
• If the instrument is linear,
– its output can be set to zero for zero input. Then a one-point calibration
defines the calibration curve that plots output versus input (next slide).
• If the linearity is unknown,
– a two-point calibration should be performed and these two points plus the
zero point plotted to ensure linearity (next slide).
• If the resulting curve is nonlinear,
– many points should be measured and plotted to obtain the calibration curve.
• If the output cannot be set to zero for zero input,
– measurements should be performed at zero and full scale for linear
instruments and at more points for nonlinear instruments.
• Calibration curves should be obtained at several expected
temperatures to determine temperature drift of the zero point and
the gain.
44
…Calibration
(a) The one-point calibration may miss nonlinearity.
(b) The two-point calibration may also miss nonlinearity.
Output
Output
Input
(a)
Input
(b)
45
Statistics
• Mean
– If we make n measurements of x, for example of
the weights of a population, we may wish to report
an estimate of them in a condensed way.
– The simplest statistic is the estimated sample mean
x

x
i
n
where i = 1, 2,…n.
46
Statistics
• Standard Deviation
– A measure of the spread of data about the mean is
the estimated sample standard deviation
s
 ( x x )
2
i
n 1
• Used with the mean for symmetric
distributions of numerical data.
47
Statistics
• Standard deviation of the mean (standard error
of the mean (SEM))
– Expresses the variability to be expected among the
means in future samples, whereas the standard
deviation describes the variability to be expected
among individuals in future samples.
sx 
s
n 1
48
Gaussian Distribution…
• The spread (distribution) of data may be
rectangular, skewed, Gaussian, or other.
• The Gaussian distribution is given by
f (X ) 
e
 ( X   ) 2 /(2 2 )
2 
where μ is the true mean and σ is the true
standard deviation of a very large number of
measurements.
49
…Gaussian Distribution
• For the normal distribution, 68% of the data lies within ±1 standard
deviation. By measuring samples and averaging, we obtain the estimated
mean , which has a smaller standard deviation sx.  is the tail probability
that xs does not differ from  by more than .
Estimated mean xs
standard deviation sx
Population standard
deviation 
Frequency




+
x
Mean
50
Poisson Probability…
• The Poisson probability density function is another type of
distribution.
– It can describe, among other things, the probability of radioactive decay
events, cells flowing through a counter, or the incidence of light
photons.
• The probability that a particular number of events K will occur
in a measurement (or during a time) having an average number
of events m is
emmK
p( K , m) 
K!
• The standard deviation of the Poisson distribution is
m
51
…Poisson Probability
• A typical Poisson distribution for m = 3.
0.2
x
x
x
x
p
0.1
x
x
0
0
1
2
3
4
5
K
52
Hypothesis testing…
• In hypothesis testing, there are two hypotheses.
– H0, the null hypothesis, is a hypothesis that assumes that the variable in
the experiment will have no effect on the result
– Ha is the alternative hypothesis that states that the variable will affect
the results.
• For any population, one of the two hypotheses must be true.
• The goal of hypothesis testing is to find out which hypothesis
is true by sampling the population.
• In reality, H0 is either true or false and we draw a conclusion
from our tests of either true or false. This leads to four
possibilities (next slide)
53
…Hypothesis testing…
• The four outcomes of hypothesis testing.
Conclusion
Real situation
H0 true
Ha true
Accept H0
Correct decision
Type II error, p = b
Reject H0
Type I error, p = a
Correct decision
54
…Hypothesis testing…
• Equivalent table of the table given in previous slide
for results relating to a condition or disease.
Test result
Has condition?
No
Yes
Negative
True negative (TN)
False negative (FN)
Postitive
False positive (FP)
True positive (TP)
55
…Hypothesis testing…
• The terms in the Table in previous slide are
useful for defining measures that describe the
proportion of, for example, a disease in a
population and the success of a test in
identifying it.
• Incidence
– is the number of cases of a disease during a stated
period, such as x cases per 1000 per year.
56
…Hypothesis testing…
• Prevalence
– the number of cases of a disease at a given time
such as y cases per 1000.
• It is all diseased persons divided by all persons.
TP  FN
Prevalence 
TN  TP  FN  FP
57
…Hypothesis testing…
• Sensitivity
– the probability of a positive test result when the disease is present.
– Among all diseased persons, it is the percent who test positive.
Sensitivit y 
TP
100 %
TP  FN
• Specificity
– the probability of a negative diagnostic result in the absence of the
disease.
– Among all normal persons, it is the percent who test negative.
TN
Specificit y 
100 %
TN  FP
58
…Hypothesis testing…
• Considering only those who test positive, positive predictive
value (PPV) is the ratio of patients who have the disease to all
who test positive.
PPV 
TP
100 %
TP  FP
• Considering only those who test negative, negative predictive
value (NPV) is the ratio of nondiseased patients to all who test
negative.
TN
NPV 
100 %
TN  FN
59
…Hypothesis testing
• The test result threshold is set to minimize
false positives and false negatives.
Normal
population
Threshold
True
negative
False
positive, p = 
Diseased
population
False
negative, p = b
True
positive
60